VideoHelp Forum




+ Reply to Thread
Results 1 to 12 of 12
  1. I'm still learning, and I love that I can shrink down blu-ray & dvd rips while keeping seemingly intact picture quality . . but with dvd's mostly, I'm always noticing a decrease in the playback/motion. The best way I can describe it is when you look at a 120hz TV compared to a 60hz TV. The 120hz has that great looking motion where you eye seems to catch every frame. It looks more 'real life' when things are moving. That's how I'd describe the mpeg2 playback I'm seeing.

    After converting in Handbrake to MKV, this seems to be lost. The motion just isn't handled as well. I especially notice it with rips of live action sitcoms. Any way to stop this from happening?
    Quote Quote  
  2. Conversion shouldn't effect that way motion looks. It might be the video isn't being de-interlaced or IVTC'd correctly. Unless a Handbrake user comes along with a clever theory, maybe try uploading a small sample of one of the DVDs you're having problems with, along with your encode of the sample for others to look at.

    If you have DVD Shrink you can open the ripped files and use it's re-author function to resave just a small section of the DVD as a single vob file.
    Quote Quote  
  3. I'll try to upload a couple of samples. It just seems to move much more fluidly in the mpeg2. Its nothing you'd notice if you just watch the MKV, but when you watch the mpeg2 after, the difference is pretty clear IMO. Like I said, similar to going from a 60hz TV to a 120hz TV.
    Quote Quote  
  4. you're probably single rate deinterlacing in handbrake 59.94i => 29.97p instead of 59.94p (I don't know if handbrake can bob deinterlace yet)
    Quote Quote  
  5. Is there any alternative to Handbrake I can test that theory with?
    Quote Quote  
  6. Originally Posted by Forty0zFreedom View Post
    Is there any alternative to Handbrake I can test that theory with?
    I'd examine the results first. If you don't know how to do that, take the suggestion earlier to upload samples here (or to MediaFire or Sendspace). 10 seconds of before and after with steady motion from the same scene will be plenty.
    Quote Quote  
  7. Alright, here are the sample files.

    MPEG2:

    http://www.mediafire.com/download.php?48g49t24g4b92ea

    MKV:

    http://www.mediafire.com/download.php?z3t9b38v62y3w6m

    I really appreciate any input. I'd love to get this right.
    Quote Quote  
  8. pdr nailed it. It's the difference in motion fluidity between the MPEG's 59.94 fields per second and the MKV's 29.97 frames per second. Doesn't bother me, though.

    You can't just reencode the DVD to a smaller size? That particular TV series was shot on video. For movies and other TV series shot on film, you won't have that particular problem.
    Quote Quote  
  9. Was the MKV sample encoded from the mpeg source you linked to? I'm just wondering why the logo at the top doesn't stay on for the same duration in both samples.

    Maybe my brain doesn't work the same way as I'm not seeing much of a difference, but could the resizing be contributing at all (given the width has been cropped but it's been resized back to 720)?
    Last edited by hello_hello; 26th Apr 2012 at 01:22.
    Quote Quote  
  10. Originally Posted by hello_hello View Post

    Maybe my brain doesn't work the same way as I'm not seeing much of a difference, but could the resizing be contributing at all (given the width has been cropped but it's been resized back to 720)?
    Primarily it's a motion samples issue. 59.94 will give you smoother results than 29.97

    How are you watching the mpg source? If you are watching on the computer , your software has to be set to bob (that's how it will look in a DVD player)

    If you are familiar with action shooter video games, when your FPS drops below ~30 it gets laggy and unplayable right? It's the same idea here: more real motion samples per second mean smoother experience

    Download this little sample video that jagabo made 24p vs 30p vs 60p. Of course real content typically isn't like this, but it's meant to clearly illustrate the issue
    https://forum.videohelp.com/images/guides/p1981421/24v30v60.avi

    The sample video @ 59.94p, is attached below , compare to the one at 29.97p





    Forty0zFreedom - What manono says above is important - this only applies to true interlaced content.

    DVD film material (e.g. Hollywood movies) will be 23.976p, and usually handbrake will inverse telecine them correctly (the filter is called "delecine" or something like that in handbrake) . For DVD sports material, soap operas, these are usually shot at 59.94i, not film rate of 23.976. Deinterlacing (single or double) is the WRONG thing to do, it will reduce the resolution and quality substantially

    Sorry I'm not up to date on all the GUIs, but what you want to look for is a bob deinterlace filter. Avisynth offers the best selection and quality, but many of the GUIs use avisynth filters in the backend. I think maybe megui , ripbot , xvid4psp should have the option to bob deinterlace (they won't usually do it automatically you have to specify it directly, otherwise single rate is the default for most software).
    Image Attached Files
    Quote Quote  
  11. pdr nailed it. It's the difference in motion fluidity between the MPEG's 59.94 fields per second and the MKV's 29.97 frames per second. Doesn't bother me, though.

    You can't just reencode the DVD to a smaller size? That particular TV series was shot on video. For movies and other TV series shot on film, you won't have that particular problem.
    I can definitely reencode but wouldn't that just give me the same mpeg2 file? Do you mean I should rip it again straight to an MKV with the right interlacing? I might be crazy, but I do notice it to a smaller extent on movies too. Maybe its my eyes playing tricks on me.

    Was the MKV sample encoded from the mpeg source you linked to? I'm just wondering why the logo at the top doesn't stay on for the same duration in both samples.
    I used a trial software to split the video. I have no idea why it added that watermark on just the MKV.

    Forty0zFreedom - What manono says above is important - this only applies to true interlaced content.

    DVD film material (e.g. Hollywood movies) will be 23.976p, and usually handbrake will inverse telecine them correctly (the filter is called "delecine" or something like that in handbrake) . For DVD sports material, soap operas, these are usually shot at 59.94i, not film rate of 23.976. Deinterlacing (single or double) is the WRONG thing to do, it will reduce the resolution and quality substantially

    Sorry I'm not up to date on all the GUIs, but what you want to look for is a bob deinterlace filter. Avisynth offers the best selection and quality, but many of the GUIs use avisynth filters in the backend. I think maybe megui , ripbot , xvid4psp should have the option to bob deinterlace (they won't usually do it automatically you have to specify it directly, otherwise single rate is the default for most software).
    This is all enlightening. Do you remember if Handbrake automatically does the correct filter on 23.976p material or is that an option you have to select? I'm just wondering if I'm not getting the right fluidity on movies too.

    But regarding the 59.94i material (If I have this right), basically I'm gonna use one of those programs instead of handbrake, select bob deintelface filter, and I can get a smaller file with that smooth motion?

    Thanks for all the help. I feel like I'm on the right track now.
    Last edited by Forty0zFreedom; 26th Apr 2012 at 15:20.
    Quote Quote  
  12. Originally Posted by Forty0zFreedom View Post
    I can definitely reencode but wouldn't that just give me the same mpeg2 file?
    Yes, and with the interlacing preserved so it'll play on a TV set with the fluidity you lost after deinterlacing it to progressive 29.97fps.
    Quote Quote  



Similar Threads

Visit our sponsor! Try DVDFab and backup Blu-rays!